The lottery is a fixture in American culture, with people spending upward of $100 billion on tickets each year. When jackpots reach hundreds of millions or even a billion dollars, a fever sweeps the nation. It’s not hard to understand why: A million-dollar prize would change many people’s lives for the better. But the fact is, that eye-popping sum comes with significant costs, both to the winner and society at large.

One major cost is that low-income people play a greater percentage of the games relative to their disposable incomes, and many studies have found that they spend more on tickets than those with higher incomes. As a result, critics charge that the lotteries are a disguised tax on those who can least afford it.

Another major cost is that states often pay out a large proportion of the sales in prizes, which reduces the percentage available for state revenue and for use on things like education. While lottery funds do provide a valuable source of revenue for state governments, they’re not as transparent as a normal tax and many consumers don’t realize the implicit taxes they’re paying when buying a ticket.

As for the actual prizes, they can be a fixed amount of cash or goods, or they may be a percentage of lottery receipts. Either way, the prize allocation is entirely based on chance and can’t be rationally justified by decision models based on expected value maximization. However, if entertainment value and other non-monetary values are factored into the utility function, a ticket purchase can be considered rational under those conditions.